Viacom International, Inc. v. Youtube, Inc.

Filing 387

REPLY BRIEF, on behalf of Appellant Viacom International, Inc., FILED. Service date 04/28/2011 by CM/ECF. [276111] [10-3270]

10-3270 IN THE UNITED STATES COURT OF APPEALS FOR THE SECOND CIRCUIT VIACOM INTERNATIONAL INC., COMEDY PARTNERS, COUNTRY MUSIC TELEVISION, INC., PARAMOUNT PICTURES CORPORATION, BLACK ENTERTAINMENT TELEVISION LLC, Plaintiffs-Appellants, v. YOUTUBE, INC., YOUTUBE, LLC, GOOGLE, INC., Defendants-Appellees. ON APPEAL FROM THE UNITED STATES DISTRICT COURT FOR THE SOUTHERN DISTRICT OF NEW YORK REPLY BRIEF FOR PLAINTIFFS-APPELLANTS Paul M. Smith William M. Hohengarten Scott B. Wilkens Matthew S. Hellman JENNER & BLOCK LLP 1099 New York Avenue, NW Washington, DC 20001 (202) 639-6000 Theodore B. Olson Matthew D. McGill GIBSON, DUNN & CRUTCHER LLP 1050 Connecticut Avenue, NW Washington, DC 20036 (202) 955-8500 Susan J. Kohlmann JENNER & BLOCK LLP 919 Third Avenue New York, NY 10022 (212) 891-1600 Stuart J. Baskin SHEARMAN & STERLING LLP 599 Lexington Avenue New York, NY 10022 (212) 848-4000 TABLE OF CONTENTS Page TABLE OF AUTHORITIES ....................................................................................iii INTRODUCTION .....................................................................................................1 ARGUMENT .............................................................................................................4 I. YouTube Built Its Business On Preventable Infringement...................4 II. Viacom—Not YouTube—Is Entitled To Summary Judgment On YouTube’s DMCA Affirmative Defense........................................8 A. YouTube’s Failure To Take Steps To Mitigate Infringement In The Face Of Its Knowledge Of Infringing Material And Awareness Of Infringing Activity Disqualifies It From The DMCA Safe Harbor .............9 1. 2. YouTube’s Disputed Evidence Relating To Affirmative Defenses Does Not Negate Its Disqualifying Awareness Of Infringing Activity Or Its Failure To Take Any Action To Stop It ...............16 3. YouTube’s Willful Blindness Satisfies Its Invented Requirement Of URL-Specific Knowledge of Infringement ...................................................................18 4. B. URL-Specific Knowledge Of Infringement Is Not Necessary To Demonstrate Disqualifying Awareness.......................................................................10 Viacom’s Evidence That YouTube Had URLSpecific Knowledge Of Infringement Is More Than Sufficient To Survive Summary Judgment ...........23 The Safe Harbor Is Unavailable To YouTube Because It Profited Directly From Infringing Activity It Could Have Controlled But Chose To Foster ...............................................24 i Table of Contents (Continued) Page 1. 2. C. III. YouTube Had The Ability To Control The Infringing Activity That Pervaded Its Network .............25 YouTube Profited Directly From The Infringing Activity It Encouraged....................................................32 YouTube’s Performance And Licensing Of UserUploaded Copyrighted Content Are Not The Type Of Storage Activities That The DMCA Immunizes From Liability .....................................................................................35 Viacom Is Entitled To Summary Judgment On Its Affirmative Claims..................................................................................................36 A. Viacom Is Entitled To Summary Judgment On Its Direct Infringement Claim ...................................................................36 B. Viacom Is Entitled To Summary Judgment On Its Grokster Claim For Intentional Facilitation Of Infringement..............................................................................37 C. Viacom Is Entitled To Summary Judgment On Its Vicarious Liability Claim .........................................................38 CONCLUSION........................................................................................................39 ii TABLE OF AUTHORITIES Cases Page(s) ALS Scan, Inc. v. RemarQ Communities, Inc., 239 F.3d 619 (4th Cir. 2001) ................................................................................19 Apollo Fuel Oil v. United States, 195 F.3d 74 (2d Cir. 1999) ...................................................................................26 Cartoon Network LP v. CSC Holdings, Inc., 536 F.3d 121 (2d Cir. 2008) .................................................................................36 Cass Cnty. Music Co. v. C.H.L.R., Inc., 88 F.3d 635 (8th Cir. 1996) ..................................................................................33 Gershwin Publ’g Corp. v. Columbia Artists Mgmt., Inc., 443 F.2d 1159 (2d Cir. 1971) ...............................................................................33 Hard Rock Café Licensing Corp. v. Concession Services, Inc., 955 F.2d 1143 (7th Cir. 1992)..............................................................................20 In re Aimster Copyright Litig., 334 F.3d 643 (7th Cir. 2003) ................................................................................19 Metro-Goldwyn-Mayer Studios Inc. v. Grokster, Ltd., 545 U.S. 913 (2004) ........................................................................................ 7, 38 Nationwide Mut. Ins. Co. v. Darden, 503 U.S. 318 (1992) ...................................................................................... 19, 28 Papelino v. Albany College of Pharm. of Union Univ., 633 F.3d 81 (2d Cir. 2011) .....................................................................................4 Perfect 10, Inc. v. CCBill LLC, 488 F.3d 1102 (9th Cir. 2007)................................................................. 33, 35, 39 Shapiro, Bernstein & Co. v. H.L. Green Co., 316 F.2d 304 (2d Cir. 1963) .................................................................................28 Tiffany (NJ) Inc. v. eBay Inc., 600 F.3d 93 (2d Cir. 2010) ........................................................................... passim iii Table of Authorities (Continued) Tiffany (NJ) Inc. v. eBay, Inc., 576 F. Supp. 2d 463 (S.D.N.Y. 2008) ..................................................................23 UMG Recordings, Inc.v. Veoh Networks Inc., 665 F. Supp. 2d 1099 (C.D. Cal. 2009)................................................................31 United States ex rel. Eisenstein v. City of New York, 129 S. Ct. 2230 (2009) .................................................................................. 26, 31 United States v. Liranzo, 944 F.2d 73 (2d. Cir. 1991) ....................................................................................2 United States v. MacPherson, 424 F.3d 183 (2d Cir. 2005) .................................................................................24 United States v. Peterson, 394 F.3d 98 (2d Cir. 2005) ...................................................................................11 United States v. Rybicki, 354 F.3d 124 (2d Cir. 2003) .................................................................................11 Statutes 17 U.S.C. § 512(c)(1)(A)(i) .....................................................................................10 17 U.S.C. § 512(c)(1)(A)(ii) ....................................................................... 10, 11, 12 17 U.S.C. § 512(c)(1)(A)(iii) ...................................................................................13 17 U.S.C. § 512(m) ..................................................................................................20 Other Authorities Black’s Law Dictionary (6th ed. 1990) ...................................................................27 H.R. Rep. No. 105-551 ............................................................................... 20, 28, 33 Justin Green, YouTube Copyright Education (remixed), Broadcasting Ourselves (Apr. 14, 2011) ......................................................................................7 S. Rep. No. 105-190......................................................................................... passim Webster’s Second New International Dictionary (1955).................................. 12, 27 iv INTRODUCTION This case involves an immense and extraordinarily successful enterprise knowingly based on the illicit use of property belonging to others. YouTube facilitated, encouraged and thrived on copyright infringement on a massive scale. In the words of the district court, a jury could conclude that the defendants “not only were generally aware of, but welcomed,” copyright infringement, and that they did so because it “enhanced defendants’ income.” SPA9. Indeed, the evidence showed that YouTube “welcomed” tens of thousands of infringing views per day. In the words of YouTube’s lead product manager, “probably 75-80% of our views come from copyrighted material.” JAII-47. YouTube asserts that its masterfully constructed and richly rewarding sponsorship of illegal activity is “legally irrelevant.” YT Br. 23 n.5. If YouTube operated a large amusement park instead of the world’s largest online mall of video clips, its position would be that, so long as it avoided specific, concrete knowledge of which of the booths it invited into the park conducted illegal activities, it would be “legally irrelevant” even if it knew that 80% of its customers patronized booths engaged in conduct such as the sale of stolen goods, dispensing alcohol to minors, and prostitution. YouTube’s theory is that even if it had the absolute right and unquestioned ability to prevent its enterprise from being put to those illegal uses, it would be free to let the unlawful conduct flourish, and profit handsomely from it. 1 But see United States v. Liranzo, 944 F.2d 73, 79 (2d. Cir. 1991) (setting forth elements of offense of criminal facilitation). YouTube’s extreme position is not a tenable interpretation of the Digital Millennium Copyright Act (“DMCA”). Its myopic view of the exclusions from Section 512(c)’s safe harbor is untethered to the statutory text, unsupported by the legislative history, and manifestly unsound as a matter of policy. It would, defendants do not deny, immunize even an avowedly piratical business that explicitly urged its users to steal copyrighted materials, and would permit such a service provider to profit from its users’ illegal activities simply by blinding itself to the real (and “welcomed”) source of its enormous profits. That is not what Congress either enacted or intended. This Court should reverse the district court’s award of summary judgment predicated on that construction for each of the following four independent reasons: 1) “Infringing activity” was at least “apparent” to YouTube, which recognized that up to 80% of its views were infringing. Yet, rather than “act expeditiously” to stop the infringement—rather than deploy the anti-infringement measures that it already had in hand— YouTube chose to do nothing except respond to URL-specific cease-and-desist demands. 2 2) Even if, as YouTube argues, “infringing activity” could not be “apparent” to a service provider unless it had URL-specific knowledge of infringement, Viacom presented evidence that unquestionably raises a triable question of fact as to whether YouTube had that required knowledge as a result of either its own explorations of the extent of infringement on YouTube or its willful blindness to that specific knowledge. 3) YouTube obtained immense financial benefits from infringing activity that it easily could have restrained, but chose to permit (“welcomed”) precisely because it enhanced YouTube’s position in the marketplace. 4) Viacom’s copyright infringement claims stem not merely from YouTube’s storage of infringing works, but also from YouTube’s own acts of copying and distribution to third parties, activities which are not encompassed by Section 512(c)’s safe harbor because they are not “by reason of storage at the direction of the user.” For any or all of these reasons, the judgment of the district court should be reversed. 3 ARGUMENT I. YouTube Built Its Business On Preventable Infringement Viacom’s evidence demonstrates that YouTube obtained its commanding position in the online video marketplace by offering what its more established competitors like Google and Microsoft would not: others’ copyrighted videos for free. YouTube purposefully facilitated rampant infringement by welcoming infringing videos onto its site, and, rather than taking reasonable and readily available steps to stop those acts of infringement, chose to reap profits from them until it was served with a URL-specific cease-and-desist demand. “YouTube disagrees,” YT Br. 22 n.5, but that “disagree[ment]” cannot justify the district court’s award of summary judgment to YouTube. Pointing to press releases and made-for-litigation factual averments, YouTube insists that, from its inception, it was a business with a “legitimate purpose and operation” that “took copyright issues seriously.” YT Br. 5-6. But this case comes to this Court on the district court’s grant of summary judgment to YouTube. It is axiomatic that, on summary judgment, this court must “construe the evidence in the light most favorable to,” and “draw all reasonable inferences and resolve all conflicts and ambiguities in favor of” Viacom as the non-moving party. Papelino v. Albany College of Pharm. of Union Univ., 633 F.3d 81, 85 n.1 (2d Cir. 2011). 4 Thus, in analyzing whether YouTube is entitled to DMCA immunity from Viacom’s claims of copyright infringement, this Court must take the following facts as true: Soon after initiating service, YouTube knew that up to 80% of the views of YouTube videos were unauthorized uses of copyrighted material. This is demonstrated by the contemporaneous words of YouTube’s lead product manager Maryrose Dunton (“the truth of the matter, is probably 75-80% of our views come from copyrighted material,” JAII-47) and YouTube founder Steve Chen (without the “obviously copyright infringing” videos YouTube would go “from 100,000 views a day down to about 20,000 views or maybe even lower,” JAII-159).1 To a significant extent, YouTube knew what copyrighted videos users were infringing. This is shown by the memorandum of YouTube founder Jawad Karim to YouTube’s board of directors stating that “blatantly illegal” clips of several Viacom properties (“South Park, MTV Cribs, Daily Show, Reno 911, Dave Chappelle”) were available on YouTube. JAII-183. 1 YouTube asserts that Chen was referring not to copyrighted videos, but socalled “stupid videos.” YT Br. 12. But Chen’s email unambiguously refers to “the copyright infringement stuff” and responds immediately to a suggestion from co-founder Jawad Karim that YouTube “just remove the obviously copyright infringing stuff.” JAII-159-160. In any event, YouTube has no explanation whatsoever for Dunton’s substantively identical admission. 5 YouTube repeatedly took deliberate action to avoid learning the URLlocation of infringing videos. For example, after allowing YouTube’s army of users to flag copyrighted content for YouTube’s attention for all of two weeks, YouTube shut down the feature out of concern that “basically if we don’t remove [the feature] we could be held liable for being served a notice.” JAII-177. Though it knew infringing videos pervaded its site, YouTube adopted a no-action policy, retaining infringing clips unless and until it received a URLspecific cease-and-desist demand from a copyright holder. As Chen explained to his co-founders: “i really don’t see what will happen. what? someone from cnn sees it? he happens to be someone with power? he happens to want to take it down right away. he gets in touch with cnn legal. 2 weeks later, we get a cease & desist letter. we take the video down.” JAII-173. YouTube easily could have curbed the infringement of its users. Viacom offered evidence that digital fingerprint filtering technology—which identifies, at the point of upload, videos matching the profile of copyrighted works supplied by content owners—was commercially available at a reasonable cost when YouTube launched. SJAI-211-12, 231-33; see generally SJAI-619-86. Yet because copyrighted works were a “major lure for [its] users,” YouTube elected to offer its “Claim Your Content” (CYC) tool, which included fingerprint filtering technology from Audible Magic, only to those “partners who enter[ed] into a revenue deal 6 with [it].” JAII-634; JAI-817; JAII-785. Google’s Sales Engineer relayed “the final verdict” about the “distribution of these tools” in an email to other Google employees stating: “CYC: should only be given to signed, non-music partners.” JAII-785; see also JAII-324-29, 353-55, 575-76, 673, 786, 791-92; SJAI-157-60; SJAIV-773-74. YouTube profited immensely from the infringing videos it purposefully chose to maintain on its site. It profited from the audience attracted by the free availability of copyrighted material, and, even more directly, it profited from the advertisements YouTube appended to infringing clips. SPA9; JAII-171, 228. In short, Viacom’s evidence demonstrates that, like another infamous content-sharing service, YouTube “unlawfully expropriated property . . . as part of the startup capital for [its] product.” Tr. of Oral Argument at 36-37, Metro-GoldwynMayer Studios Inc. v. Grokster, Ltd., 545 U.S. 913 (2004) (No. 04-480). That the YouTube of today is increasingly effective in filtering copyrighted material and sends infringing users to “copyright school” is of no moment, except perhaps to demonstrate what YouTube could have been doing during the time period that is at issue in this lawsuit, from YouTube’s launch in 2005 to May 2008. Justin Green, YouTube Copyright Education (remixed), Broadcasting Ourselves (Apr. 14, 2011), http://youtube-global.blogspot.com/2011/04/youtube-copyrighteducation-remixed.html. During that time of meteoric growth, YouTube repeat- 7 edly refused to take readily available steps to restrain the rampant and blatant infringement of which it indisputably had knowledge and the ability to curtail. It did so because, in founder Chen’s words, “we need views,” JAII-162, and as the district court found, it was copyrighted material—not bulldogs on skateboards or other “stupid videos”—that was “attractive to users.” SPA9; YT Br. 12. The central question in this case is whether that conduct was lawful. This Court’s answer is important not only because it will guide YouTube’s future conduct, but also because it will guide that of future YouTubes. II. Viacom—Not YouTube—Is Entitled To Summary Judgment On YouTube’s DMCA Affirmative Defense YouTube’s sweeping assertion of immunity cannot be reconciled with the text of Section 512(c). To shoehorn a business built primarily on infringement into a safe harbor intended for innocent service providers who perform certain storage functions, YouTube must contort the statutory language such that “aware of facts and circumstances from which infringing activity is apparent” means possessing particularized knowledge that material at a specific URL address is infringing, and, contrary to the common law, excludes knowledge that would have been obtained but for the provider’s willful blindness. YT Br. 29-34, 37-40. “Right and ability to control [infringing] activity” must be confined to meaning the right to direct the primary infringer as a principal directs an agent, and whether one obtains a “financial benefit directly attributable to the infringing activity” depends on the strength 8 of its claim to a “legitimate business model.” Id. at 58-69, 73-77. And “storage at the direction of a user” must encompass any act of “distribution” of the stored materials that the provider might wish to undertake. Id. at 77-81. The statutory language cannot bear these interpretations—a fact that is amply supported by legislative history demonstrating that Congress intended the DMCA’s safe harbors to encompass only those service providers who respect the copyrights of others. It does not immunize service providers who intentionally facilitate and profit from their users’ acts of theft. A. YouTube’s Failure To Take Steps To Mitigate Infringement In The Face Of Its Knowledge Of Infringing Material And Awareness Of Infringing Activity Disqualifies It From The DMCA Safe Harbor In its opening brief, Viacom showed that the district court erred when it held that Section 512(c)’s knowledge and awareness exclusions both require URLspecific knowledge of infringement, even though the awareness exclusion applies “in the absence of such actual knowledge.” YouTube’s latest efforts to justify this counter-textual construction of the statute fail as well. But even if that construction were correct—and it is not—the district court failed even to address Viacom’s evidence that YouTube willfully blinded itself to and otherwise possessed precisely the URL-specific knowledge it claims is necessary. That evidence precludes summary judgment even on YouTube’s view of the law. 9 1. URL-Specific Knowledge Of Infringement Is Not Necessary To Demonstrate Disqualifying Awareness The district court’s construction of Section 512(c)’s awareness exclusion is erroneous because, in holding that the knowledge and awareness exclusions both require proof of URL-specific knowledge, it accords no independent meaning to the awareness exclusion; it excludes no service provider not already excluded by the knowledge provision. In an effort to salvage the district court’s result, YouTube proposes (at 32) that the awareness provision has independent force because it enables a copyright holder to prove URL-specific knowledge of the legal conclusion of infringement objectively, whereas under the knowledge provision a rights holder would have to prove that the service provider has subjective knowledge of the legal conclusion that material at a particular URL is infringing. Thus, YouTube argues, the knowledge and awareness provisions provide copyright owners with two possible means of proving “the same thing.” YT Br. 32. The inescapable flaw in this argument is that the knowledge and awareness provisions do not require proof of “the same thing.” The knowledge provision requires proof “that the material or an activity using the material . . . is infringing.” 17 U.S.C. § 512(c)(1)(A)(i) (emphasis added). The awareness provision requires proof of “infringing activity.” Id. § 512(c)(1)(A)(ii) (emphasis added). YouTube suggests that the awareness provision’s term “infringing activity” means the same thing as the knowledge provision’s term “an activity using the [in10 fringing] material,” and that, therefore, both require “knowledge of specific instances of infringement.” YT Br. 32 n.8. But “[w]hen Congress uses particular language in one section of a statute and different language in another, we presume its word choice was intentional.” United States v. Peterson, 394 F.3d 98, 107 (2d Cir. 2005). In the DMCA, Congress’s use of the definite article “the” in the knowledge provision’s phrase “the material or an activity using the material,” suggests that Congress was referring there to knowledge pertaining to specific, previously-defined “material”—to wit, the “material that resides on a system.” See United States v. Rybicki, 354 F.3d 124, 137-138 & n.12 (2d Cir. 2003) (en banc). On the other hand, the absence of the definite article in the awareness provision’s phrase “infringing activity”—or the “such” construction employed in that provision’s reference to the “actual knowledge” in the knowledge provision—suggests that Congress was referring back not to the previously-defined “activity,” but instead to any “infringing activity.”2 See 17 U.S.C. § 512(c)(1)(A)(ii) (“in the absence of such actual knowledge”). 2 YouTube points to legislative history to argue that “the term ‘activity’” was included in Section 512(c) “to make clear that the statute covered infringing ‘activity using the material on the system or network.’” YT Br. 32 n.8 (quoting S. Rep. No. 105-190, at 44 (1998)). But the legislative history YouTube invokes discusses the knowledge provision, not the awareness provision, and therefore has no bearing on the proper interpretation of the awareness provision’s term “infringing activity.” 11 Indeed, it is just plain English that one can be aware of “infringing activity” without having specific knowledge of what copyrighted material has been infringed. If one’s child shouts from his computer, “I’m using Grokster to download a bunch of copyrighted songs without paying for them,” there can be no doubt that the parent has knowledge of his child’s “infringing activity,” even if he does not know which copyrights his child has infringed. On the other hand, YouTube’s construction of “infringing activity,” which requires “knowledge of specific instances of infringement,” (YT Br. 32 n.8) cannot be reconciled with the first clause of the awareness exclusion, which states that it applies “in the absence of such actual knowledge,” which is to say exactly the knowledge that “the material . . . is infringing” that YouTube says is required under its interpretation of the awareness provision. Contrary to YouTube’s assertions, according “infringing activity” its natural meaning would not close the safe harbor to a service provider upon a mere showing of “generalized awareness that unidentified infringing material may be somewhere on the service provider’s system.” YT Br. 3. The statute requires awareness of “facts [and] circumstances” that make infringing activity “apparent,” i.e., “[c]lear or manifest to the understanding; plain; evident; obvious; palpable.” 17 U.S.C. § 512(c)(1)(A)(ii); Webster’s Second New International Dictionary 129 (1955). 12 The assertedly “common knowledge” that Internet users sometimes infringe copyrights is not sufficient to make it “apparent” that infringing activity is occurring on a site. For infringing activity to be “apparent” to a service provider, it must be aware that substantial infringement is actually and regularly occurring on its site—not merely that sporadic acts of infringement might be, or likely are, occurring. Here, YouTube not only was aware of rampant and ongoing infringement, it quantified the infringing activity as constituting up to 80% of its views. See, e.g., JAII-159-60. Given the centrality of copyright infringement to YouTube’s business model, there can be no doubt that that infringing activity was at least “apparent” to YouTube. But even where the infringing activity is as apparent to a service provider as it was to YouTube, that service provider still may enjoy the protections of the safe harbor if it “acts expeditiously to remove . . . the material.” 17 U.S.C. § 512(c)(1)(A)(iii). The scope of that obligation takes account of context, requiring only that the service provider take reasonable steps to remove the infringing material. YouTube claims this test is “conjured . . . out of thin air,” YT Br. 31, but it was this Court in Tiffany (NJ) Inc. v. eBay Inc. that recognized that contributory infringement liability could turn on a provider’s implementation of antiinfringement features “as such measures became technologically feasible and reasonably available.” 600 F.3d 93, 100 (2d Cir. 2010). If a service provider aware 13 of infringing activity has no technologically feasible or reasonably available means to restrain the infringing activity, then it may not be ousted from the safe harbor. The obligation on a service provider aware of infringing activity is to do what it can—to “act expeditiously”—not to do the impossible. YouTube’s exclusion from the safe harbor thus stems not simply from its failure to remove infringing material, but from its failure to even attempt to do so, despite its access to a number of feasible and commercially reasonable steps that would have achieved this end. Most obviously, YouTube could have continued community flagging and extended digital fingerprint filtering beyond its content partners. In addition, YouTube could have used keyword searching to root out infringement, another technology that it deployed selectively for content partners. SJAI-205, 222-23, 226. Beyond that, YouTube could have removed the specific infringing videos it was aware of—for example, the “well-known shows,” all copyrighted, that Karim flagged for the Board. JAII-183. It similarly could have relied on its employees to flag any instances of apparent infringement they encountered, and designated certain employees to review the most popular clips and remove those that were obviously infringing. If, after applying these and other commercially reasonable efforts, infringing videos remained, YouTube reasonably could maintain that it had “act[ed] expeditiously to remove” the infringing materials, albeit not entirely successfully. But 14 YouTube undertook none of those efforts, and it is that refusal to do that which it easily could have done to prevent the infringement that excludes YouTube from the safe harbor. Indeed, if it were otherwise—if, as YouTube argues, its obligation to act was triggered only by URL-specific knowledge of the type found in takedown notices—then all of its belated efforts to “combat copyright abuse”— Content ID digital fingerprinting, the new “copyright school,” and the like, all assertedly undertaken in “good faith”—would be mere gratuitous gestures that it could withdraw as soon as its interpretation of the DMCA is affirmed. Neither the statute nor sound policy suggests any reason to give service providers aware of rampant infringement license to do nothing at all when there is much they can readily do.3 Indeed, such a reading of the DMCA would make a mockery of the “cooperat[ion]” and “copyright holders and service providers acting together” that even YouTube’s amici agree the DMCA was intended to foster. 3 Section 512(m) certainly does not provide service providers such an excuse. Requiring a service provider to cure ongoing infringement of which it is aware no more establishes “monitoring” or “affirmatively seeking facts indicating infringing activity” as a “condition [to] the applicability” of the safe harbor—the limited proscription of § 512(m)—than does requiring affirmative action in response to a take-down notice. 15 Br. for amicus curiae Michael Carrier at 2, 14; see also Br. for amici curiae eBay Inc. et al. at 5, 9.4 2. YouTube’s Disputed Evidence Relating To Affirmative Defenses Does Not Negate Its Disqualifying Awareness Of Infringing Activity Or Its Failure To Take Any Action To Stop It Though it was not a basis for the decision below, YouTube strains to show that, if it had made an effort to remove clips that infringed Viacom’s copyrights, it might have been difficult to distinguish thousands of unauthorized clips from those few that Viacom authorized. This, YouTube claims, would have rendered it helpless to identify and remove infringing videos and accordingly precludes exclusion from the safe harbor for failing to act to stop those infringing activities. At the threshold, this argument fails as a basis for summary judgment because it is based on YouTube’s tendentious and disputed view of the facts. For example, YouTube quotes a Viacom employee’s characterization of the uploaded clips as a “boatload,” but it fails to note that the document it cites identifies the supposed “boatload” as seven clips. JAIV-260-61. YouTube similarly accuses 4 So, too, would an interpretation that requires service providers only to respond to DMCA take-down notices. See Br. for amici eBay et al. at 7-9. That would hardly constitute a system of shared responsibility and it could not be effective to protect copyrights in any event. More to the point, however, the interpretation is foreclosed by the structure of Section 512, which “does not require use of the notice and take-down procedure.” S. Rep. No. 105-190, at 45. 16 Viacom of using “dozens of obscure accounts that bore no obvious link to Viacom,” YT Br. 15-16, but fails to identify a single account used by Viacom or its marketing agents of which YouTube was actually unaware. And, in fact, on the occasions when Viacom did place clips on YouTube, they were generally uploaded in cooperation with—indeed, encouraged by—YouTube, with usernames which it specifically established for Viacom for that purpose. SJAIV-919-21. In any event, any potential difficulty that YouTube might have faced in determining whether one of the clips in suit is authorized or subject to some other affirmative defense such as fair use is completely beside the point. YouTube nowhere contends that it believed that all of the clips in suit, or even a substantial portion of them, are subject to such defenses. Its asserted difficulty with respect to a relatively few clips does nothing to diminish its knowledge of the infringing activity that pervaded YouTube. Nor does it relieve YouTube of its obligation, as a service provider with knowledge of infringing activities, to do what it feasibly can do to stop the infringement. If YouTube had deployed just the filtering technologies it had in hand, those technologies would have automatically identified the authorized clips and given Viacom the choice as to whether they should remain on YouTube or be removed, completely alleviating this supposedly vexing difficulty. Tellingly, YouTube complains of no difficulty identifying authorized clips now 17 that it offers its Content ID filtering technology to all copyright holders rather than just those agreeing to YouTube’s licensing terms. 3. YouTube’s Willful Blindness Satisfies Its Invented Requirement Of URL-Specific Knowledge of Infringement Even if, contrary to its plain language, the awareness exclusion requires URL-specific knowledge of infringement, the district court’s award of summary judgment still must be reversed because Viacom has adduced evidence sufficient at least to raise a triable question of fact as to whether YouTube was willfully blind to that knowledge, and centuries-old common-law doctrine—applicable in copyright law as elsewhere—teaches that such willful blindness is the legal equivalent of knowledge. As this Court recently put it, “[a] service provider is not . . . permitted willful blindness”; “[w]hen it has reason to suspect that users of its service are infringing . . . it may not shield itself from learning of the particular infringing transactions by looking the other way.” Tiffany, 600 F.3d at 109. YouTube’s principal response is that the DMCA categorically excludes the doctrine of willful blindness as a theory of liability, either by including an “objective” awareness provision, or through Section 512(m)’s provision to protect users’ privacy. But these arguments are premised on a fundamental misunderstanding of the doctrine of willful blindness. Willful blindness is not a distinct mental state, like knowledge and disqualifying awareness, that Congress would be expected to identify specifically. Instead, 18 “[w]illful blindness is knowledge . . . in copyright law . . . as it is in the law generally.” In re Aimster Copyright Litig., 334 F.3d 643, 650 (7th Cir. 2003) (emphasis added), cited with approval in Arista Records LLC v. Doe 3, 604 F.3d 110, 118 (2d Cir. 2010); see also Tiffany, 600 F.3d at 110 n.16 (“The principle that willful blindness is tantamount to knowledge is hardly novel.”). Congress is deemed to have embraced this background understanding unless doing so “would thwart the congressional design or lead to absurd results.” See Nationwide Mut. Ins. Co. v. Darden, 503 U.S. 318, 323 (1992). It is YouTube’s interpretation of Section 512(m) as abolishing the doctrine of willful blindness that “would thwart the congressional design” of a safe harbor limited to “innocent service providers” that “disappears at the moment the service provider loses its innocence.” ALS Scan, Inc. v. RemarQ Communities, Inc., 239 F.3d 619, 625 (4th Cir. 2001). And it leads to the impossibly “absurd result[]” of allowing facilitators of infringement to escape liability through the artifice of willful blindness—a roadmap to intentional piracy. Such an interpretation of Section 512(m) could be adopted only if it were clearly compelled by the text of the statute; but the statutory text also refutes YouTube’s construction. Section 512(m), captioned “Protection of Privacy,” protects the privacy of users’ stored materials by providing that Section 512 shall not be construed to “condition the applicability” of the DMCA’s safe harbors on “a ser- 19 vice provider monitoring its service or affirmatively seeking facts indicating infringing activity.” 17 U.S.C. § 512(m). Imposing an obligation to restrain infringement upon a showing of willful blindness does not condition the applicability of the safe harbor on the provider undertaking monitoring or similar activities. Indeed, courts have explicitly distinguished the willful blindness doctrine from an affirmative duty to seek facts out. In Hard Rock Café Licensing Corp. v. Concession Services, Inc., 955 F.2d 1143 (7th Cir. 1992), for example, the Court specifically recognized that the law imposed “no affirmative duty [on the defendant] to take precautions against the sale of counterfeits.” Id. at 1149. Yet the Court emphasized that the defendant could nonetheless face liability if the facts demonstrated its willful blindness to such sales. Id. The legislative history of Section 512(m) precisely tracks this distinction; a provider has “no obligation . . . to seek out such red flags,” but “[o]nce a provider becomes aware of a red flag, however, it ceases to qualify for the exemption.” H.R. Rep. No. 105-551, pt. 1, at 25 (1998). Without its categorical legal argument, YouTube’s factual disputes with Viacom’s evidence that YouTube was willfully blind to URL-specific knowledge of infringement are insufficient to salvage the district court’s award of summary judgment. To be sure, YouTube has post hoc rationalizations for its termination of its copyright community flagging and its selective deployment of filtering technol- 20 ogy. But the contemporaneous documentary record (which the district court did not appear to consider) unmistakably demonstrates that YouTube purposefully avoided knowledge that, even on YouTube’s misreading of the statute, would impose an obligation on it to stanch the tide of infringement. For example, YouTube states that it discontinued community flagging for “legal reasons and because it was ineffective.” YT Br. 40-41. But the “legal reasons” are clear; in an unambiguous internal email, YouTube co-founder Chad Hurley warns that “if we don’t remove [the feature] we could be held liable for being served a notice.” JAII-177. A jury could accordingly conclude that YouTube killed its community flagging feature because it worried that such flags would amount to “notice” that particular videos were infringing. YouTube thus closed its eyes to precisely the knowledge it claims the knowledge and awareness exclusions both require.5 YouTube’s disputes with Viacom’s evidence that it offered Audible Magic and similar tools “only . . . to partners who enter[ed] into a revenue deal with [it]” are even more flimsy. JAI-817; JAII-675-77. In an attempt to deflect attention 5 There is no evidence, on the other hand, that even remotely suggests that YouTube terminated community flagging for copyright because it was “ineffective.” YouTube now argues (at 41) that some personal videos were erroneously flagged as copyrighted, but the possibility of similar errors does not make YouTube’s highly successful community flagging for pornography and hate speech “ineffective.” 21 from this policy, YouTube claims that it “license[d] Audible Magic’s audiofingerprinting system” in October 2006. But YouTube does not—because it cannot—claim that it actually used the system at that time for copyright owners like Viacom who declined to give YouTube a license. YT Br. 42, 70. Similarly, although it contends that it “started building Content ID,” its own proprietary filtering system, shortly thereafter, id. at 42-43, YouTube did not implement digital fingerprinting to prevent the infringement of Viacom’s copyrighted works until May 2008—and Viacom does not seek damages beyond that point. SJAII-322.6 Viacom’s evidence that YouTube purposefully evaded URL-specific knowledge of infringement distinguishes this case from Tiffany. But perhaps an even more telling distinction is found in the efforts undertaken by eBay to root out infringement. While YouTube did nothing but respond to DMCA-compliant takedown notices, this Court acknowledged that eBay, despite a “limited” “ability to determine whether a particular listing was for counterfeit goods,” Tiffany, 600 F.3d at 98, spent “as much as $20 million each year on tools to promote trust and safety on its website.” Id. eBay employed “4,000 employees devoted to trust and safety 6 YouTube’s alternative suggestion that its “hashing” technology was a legitimate substitute for digital fingerprint filtering is meritless. Hashing is a substantially inferior technology that has no effect if the second video clip differs from the first in even the slightest degree. Indeed, it either was not applied or did not work to identify thousands of clips identical to those Viacom identified in its 2007 takedown notice. SJAIII-553. 22 issues,” and performed extensive keyword searches to attempt to identify infringing materials. Id. eBay also instituted a “fraud engine” which used “13,000 different search rules” to “monitor . . . and flag or remove [counterfeit] listings.” Tiffany (NJ) Inc. v. eBay, Inc., 576 F. Supp. 2d 463, 477 (S.D.N.Y. 2008); see Tiffany, 600 F.3d at 98-99. The fraud engine flagged for further review thousands of listings daily that “contained obvious indicia of infringing or otherwise fraudulent activity.” Tiffany, 576 F. Supp. 2d at 477. This was all in addition to eBay’s own takedown notice system, which enabled rights owners to search the site “automatically for particular listings every day.” Id. at 478; see Tiffany, 600 F.3d at 99. In short, “eBay consistently took steps to improve its technology and develop anti-fraud measures as such measures became technologically feasible and reasonably available.” Tiffany, 600 F.3d at 100. The dramatic disparity between eBay’s responsible conduct and YouTube’s “welcom[ing]” of copyright infringement, its refusal to employ even the antiinfringement technologies it had in hand, and its intentional profiting from that infringement amply supports a finding of willful blindness here. 4. Viacom’s Evidence That YouTube Had URL-Specific Knowledge Of Infringement Is More Than Sufficient To Survive Summary Judgment YouTube ignores Viacom’s evidence that YouTube had actual knowledge of specific instances of infringement, summarily asserting that the documents on 23 which Viacom relies do not refer to specific clips. YT Br. 53. But YouTube offers no authority for the proposition that such detail is necessary; to the contrary, knowledge “can often be proved through circumstantial evidence and the reasonable inferences drawn therefrom.” United States v. MacPherson, 424 F.3d 183, 189 (2d Cir. 2005). Viacom’s evidence permits just such an inference. For example, Dunton’s review of YouTube’s top videos, through which she determined that 70 percent were copyrighted, necessarily involved identifying the copyrighted videos and necessarily gave Dunton access to the URL addresses where the videos were posted. JAI-857. Similarly, when Karim identified “blatantly illegal” videos, he must have done so by finding those specific clips on YouTube’s site and determining they were unlikely to be authorized or otherwise permissible uses. JAII-183. YouTube has not indisputably established that it lacked URL-specific knowledge of infringement as to all of the 62,637 Viacom clips in suit. B. The Safe Harbor Is Unavailable To YouTube Because It Profited Directly From Infringing Activity It Could Have Controlled But Chose To Foster The statute excludes YouTube from the safe harbor for the independent reason that it received a direct financial benefit from activity it had the right and ability to control. Certainly, YouTube had the ability to “control” infringing activity under any ordinary meaning of that term; its own brief lauds its “technological 24 measures” “to deter users from uploading unauthorized material.” YT Br. 7-8. Its contrary argument rests on its ascription of a definition to the verb “control” that is rejected even by the scant authorities YouTube cites in its brief. And it is equally clear that when YouTube decided to use others’ copyrighted works as its “major lure” for users and appended revenue-generating advertisements to infringing clips, it received financial benefits that are directly attributable to the infringing activity it fostered. YouTube’s “legitimate business” test has no mooring whatsoever in the text of the statute and cannot protect YouTube’s early conduct in any event. 1. YouTube Had The Ability To Control The Infringing Activity That Pervaded Its Network YouTube offers no meaningful defense of the ground upon which it prevailed below—to wit, that the “ability to control [infringing] activity” requires URL-specific knowledge of infringement. SPA28-29. YouTube just asserts that the control inquiry “focuses on particular instances of alleged infringing activity” and thus requires “knowledge of that activity” or “antecedent awareness of the individual materials or activities that are alleged to be infringing.” YT Br. 60. But the assertion is illogical: If a service provider exercised such effective control over infringing activity that its site was entirely free of infringement, the absence of infringement (and thus of knowledge of that infringement) would not demonstrate an inability to control it; it would (amply) demonstrate the opposite. 25 The assertion also defies the statutory text: If “knowledge” or “awareness” of infringing activity is required to establish “ability to control,” Section 512(c)(1)(B) excludes no one from the safe harbor not already excluded by the knowledge and awareness provisions of Section 512(c)(1)(A), making Section 512(c)(1)(B) superfluous. But see United States ex rel. Eisenstein v. City of New York, 129 S. Ct. 2230, 2234 (2009) (“well-established principles of statutory interpretation . . . require statutes to be construed in a manner that gives effect to all of their provisions”). YouTube says this is “wrong,” and, despite having just argued that knowledge or awareness is “require[d],” purports to give two examples of occasions in which control may be present, without knowledge of infringing activity. It first claims that a service provider can direct a user to engage in infringement while “acting as the principal in an agency relationship,” “even though it may not have knowledge.” YT Br. 60. But as a general matter, “the agent’s knowledge is imputed to the principal.” See, e.g., Apollo Fuel Oil v. United States, 195 F.3d 74, 76 (2d Cir. 1999) (per curiam). Indeed, a corporation’s knowledge is always predicated on the knowledge of its agents. A principal would accordingly never provide an example of a service provider who would be excluded from the safe harbor by Section 512(c)(1)(B), but not Section 512(c)(1)(A). 26 YouTube’s second example limits “ability to control” to circumstances where a service provider “directs or exercises controlling influence over the primary infringer.” But this definition does not accord with the natural meaning of the term “control.” YT Br. 60 (citing Black’s Law Dictionary (6th ed. 1990) (“Black’s”)). For example, a property owner can control trespassing with a fence, though it neither knows of the would-be trespassers nor directs their conduct. Indeed, that one can “control” an activity without managing its perpetrators is demonstrated by the very dictionary YouTube cites, Black’s, which defines the verb “control” to mean “[t]o regulate; restrain.” Black’s at 329; see also Webster’s Second New International Dictionary 580 (defining verb “to control” as “to exercise restraining or directing influence over”). The plain meaning of “control” thus extends to actions, such as filtering, that limit the conduct sought to be restrained, even if they fail to extinguish it. YouTube’s only apparent guideline in interpreting “right and ability to control” is that it must differ from the identically framed standard applied under the common law. To justify this unusual method of statutory interpretation, YouTube reads a comment in the Senate Report that Congress was not “embark[ing] upon a wholesale clarification” of contributory infringement and vicarious liability, S. Rep. No. 105-190, at 19, as an indication that Congress abandoned its earlier view that “the ‘right and ability to control’ language . . . codifies the second ele- 27 ment of vicarious liability . . . [and] is intended to preserve existing case law . . . ” H.R. Rep. No. 105-551, pt. 1, at 26; YT Br. 59-60. But the passage YouTube quotes nowhere indicates that Congress desired to alter, much less “abandon[],” previously expressed intent. YT Br. 59. It merely confirms that Congress decided not to statutorily clarify the doctrines of contributory infringement and vicarious liability in a way that would indirectly influence the liability of service providers, but instead “to leave current law in its evolving state and . . . create a series of ‘safe harbors,’ for certain common activities of service providers.” S. Rep. No. 105190, at 19. This decision does not affect this Court’s obligation to read terms with an established common-law meaning in light of that meaning. See Nationwide Mut. Ins. Co., 503 U.S. at 323. And under the common law, vicarious liability was available “even in the absence of actual knowledge that the copyright monopoly is being impaired.” Shapiro, Bernstein & Co. v. H.L. Green Co., 316 F.2d 304, 307 (2d Cir. 1963).7 7 The statement in the House Report that the DMCA would make the “current criteria for finding contributory or vicarious liability . . . clearer and somewhat more difficult to satisfy,” H.R. Rep. No. 105-551, pt. 1, at 13, is irrelevant to the “right and ability to control” standard. The same report separately addresses the elements of vicarious liability—“direct financial benefit” and “right and ability to control”—and explains that the DMCA “modifies and clarifies . . . the financial benefit element of vicarious liability,” while “[t]he ‘right and ability to [Footnote continued on next page] 28 YouTube finally retreats to arguing the negative: whatever the “ability to control” means, it may not be established by “the provider’s ability to exercise generalized control over its system.” YT Br. 61 (citing California district court cases). This is a counter-intuitive proposition, but it is an attack on a straw man in any event. Section 512(c)(1)(B)’s exclusion does not require an ability to control a system, but instead the right and ability to control the infringing activities from which the provider derives financial benefits. A service provider’s “generalized control over its system” of course does not itself remove the provider from the safe harbor. But there can be no doubt whatsoever that the massive amounts of infringing activities from which YouTube drew its lucrative financial benefits were activities it had the ability to restrain. It just chose not to do so, instead withholding filtering technology from content providers who declined YouTube’s licensing terms. YouTube attempts to justify this policy with a conclusory allegation that YouTube had “legitimate technological concerns about using Audible Magic,” because Audible Magic supposedly “had no experience” in the specific context of television shows and movies and was “often unreliable.” YT Br. 72-73. That view of the facts is, to say the least, disputed, and is obviously contradicted by You- [Footnote continued from previous page] control’ language . . . codifies the second element of vicarious liability.” Id. at 25. See infra pp. 33-34. 29 Tube’s willingness to use digital fingerprint filtering for those who agreed to give it a license. In any event, YouTube never disputes that Audible Magic would in fact have identified many, if not most, infringing videos. And YouTube even touts its own Content ID technology as “more effective for copyright owners than any existing technology”—yet declined to use it for Viacom for six months after its October 2007 launch. Id. at 73. In short, YouTube’s evidence demonstrates that YouTube lacked the desire to control infringement of Viacom’s copyrights, not the ability. Section 512(m) does not alter this conclusion. Simply recognizing that those with the ability to filter out infringing materials in the upload process have the ability to control infringing activity would not make the deployment of such technology a “condition [on] the applicability” of the safe harbor. Instead, companies declining to exercise their ability to control infringing activity still could claim the safe harbor so long as they did not receive financial benefits directly attributable to the activity and act expeditiously to stop the infringing activity of which they are aware. On the other hand, by excluding from the definition of “ability to control,” an ability to do anything that possibly could be construed as “affirmatively seeking facts” or “monitoring,” YouTube’s construction of Section 512(m) makes the control-and-financial-benefit exclusion of Section 512(c)(1)(B) completely ineffectual; it would include only those already possessing URL-specific knowledge and there- 30 fore capture no one not already captured by the knowledge-or-awareness exclusion of Section 512(c)(1)(A). That is an untenable construction of the statute. See Eisenstein, 129 S. Ct. at 2234. And inasmuch as it would immunize service providers that erect a business upon and intentionally profit from infringing acts that they easily could control, is utterly unjustifiable as a matter of policy.8 Finally, YouTube’s oft-asserted inability to identify authorized copyrighted material does not undermine its ability to control infringing activity. Even if YouTube might have a legitimate question whether a certain copyrighted video was authorized, it nonetheless had the ability to control or restrain the infringing activity which it could readily identify. And, indeed, YouTube does not argue that it has any difficulty distinguishing authorized copyrighted materials from unauthorized materials with its Content ID filtering system, and it had no similar difficulty with Audible Magic. 8 UMG Recordings, Inc.v. Veoh Networks Inc., 665 F. Supp. 2d 1099 (C.D. Cal. 2009) (“UMG II”), which YouTube invokes as if it were binding precedent (YT Br. 62), is incorrectly decided on this point. It, like YouTube, fails to recognize that if the safe harbor is potentially available to a service provider in the absence of monitoring or affirmative investigation of infringement, then such monitoring or investigation cannot be a “condition [on] the applicability” of the safe harbor. 31 2. YouTube Profited Directly From The Infringing Activity It Encouraged On the second aspect of Section 512(c)(1)(B), YouTube once again runs away from the text of the statute—“a financial benefit directly attributable to the infringing activity”—and urges application of what it calls a “narrower, more practical financial-benefit test,” purportedly grounded in the legislative history, that asks whether the service provider is “conducting a legitimate business.” YT Br. 74. Beyond the fact that it has no connection whatsoever to the statutory text, YouTube’s test makes no sense. Even a “legitimate business” could operate a side business that profits from others’ act of infringement. If Google had bought Grokster rather than YouTube, it could hardly be disputed that Google was obtaining financial benefits from the infringing activity, even if Google otherwise were a “legitimate business.” Not surprisingly, then, the snippet of legislative history on which YouTube relies, YT Br. 74-75, does not support its test. Instead, it merely establishes that businesses which are both legitimate and receive the same revenues regardless of the availability of infringing material do not receive a prohibited direct financial benefit. S. Rep. No. 105-190, at 44-45. Indeed, the very next sentence after the excerpt YouTube quotes—which it unsurprisingly omits—provides that a “financial benefit directly attributable to the infringing activity” has occurred if an otherwise “legitimate” service provider ac32 cepts a fee “where the value of the service lies in providing access to infringing material.” S. Rep. No. 105-190, at 44-45. This test is similar to the common law vicarious liability inquiry, which asks “whether the infringing activity constitutes a draw for subscribers, not just an added benefit.” Perfect 10, Inc. v. CCBill LLC, 488 F.3d 1102, 1117 (9th Cir. 2007); see also Ellison v. Robertson, 357 F.3d 1072, 1079 (9th Cir. 2004) (“The essential aspect of the ‘direct financial benefit’ inquiry is whether there is a causal relationship between the infringing activity and any financial benefit a defendant reaps . . .”). YouTube substantially exaggerates the extent to which the DMCA’s financial benefit standard can be read to depart from that common law standard. To be sure, the statute applies only to “financial benefit[s] directly attributable to the infringing activity”—a modifier applied in most common law vicarious liability cases, but not all. Compare Gershwin Publ’g Corp. v. Columbia Artists Mgmt., Inc., 443 F.2d 1159, 1162 (2d Cir. 1971) (requiring “direct financial interest” in infringing activity), with Cass Cnty. Music Co. v. C.H.L.R., Inc., 88 F.3d 635, 638 (8th Cir. 1996) (imposing vicarious liability on defendant who “had a financial interest in the use of the copyrighted songs”). By requiring adherence to the traditional common law standard—direct financial benefit—and disavowing those cases that depart from it, the statutory text “modifies and clarifies” the thenexisting common law standard. H.R. Rep. No. 105-551, pt. 1, at 25. That is a far 33 cry from the “reject[ion]” of the common law that YouTube has imagined and attributed to Congress. YT Br. 60. There can be no doubt that YouTube failed the test that Congress actually enacted. Viacom’s evidence demonstrates that, in the period in suit, users visited YouTube overwhelmingly to view copyrighted materials for free; by YouTube’s own estimation, 75-80% of its views were infringing. Inasmuch as its advertising revenues depended directly on the size of its audience, YouTube’s “enhanced” revenues are directly attributable to the infringing activity that “attract[ed]” the ever-larger audiences. SPA9. Indeed, Google was well aware that YouTube had built its business on infringement—leading to YouTube’s defeat of Google Video in the marketplace—yet purchased YouTube to capture the audience drawn by the infringing videos. JAI-395-96; JAII-445-46. If that were not enough, until January 2007, YouTube ran advertisements directly on the “watch pages” where it publicly displayed and performed infringing videos. JAI-310-12; JAI-792. The revenue generated from those advertisements is “directly attributable” to the infringing activity under any conceivable reading of that phrase. So, too, with the ads run on search pages displaying the results from searches for infringing content, which continued even after the watch page ads were removed once YouTube recognized that collecting advertising revenue from infringing watch pages was indefensible. JAI-312-16. 34 Though YouTube may quibble over the amount, there can be no serious dispute that YouTube obtained revenue that was “directly attributable” to the infringing activity pervading its site. Because YouTube also had the right and ability to control that infringement, the award of summary judgment to YouTube cannot stand. C. YouTube’s Performance And Licensing Of User-Uploaded Copyrighted Content Are Not The Type Of Storage Activities That The DMCA Immunizes From Liability YouTube’s expansive definition of storage as including any actions that make users’ stored materials “more accessible to the public” would protect virtually anything a service provider does to a stored work. YT Br. 77. Within this broad category, YouTube would include “reproduction, distribution, and display” of copyrighted works. Id. at 79. Whatever conduct may be protected by Section 512(c) as sufficiently related to “storage at the direction of the user,” YouTube’s manipulation and distribution of infringing videos is beyond its pale. Most obviously, YouTube’s sublicensing to third parties of copyrighted works, see JAIII-500, does not cause infringement by reason of storage, but by reason of sublicensing. Cf. Perfect 10, 488 F.3d at 1117 (Section 512(d) safe harbor unavailable where the plaintiff “does not claim that [the defendant] infringed its copyrights by providing a hyperlink,” but through “performance of other business services”). Moreover, YouTube’s primary busi- 35 ness is to perform, distribute, and promote stored videos, and to solicit revenues for advertising connected with those videos; the storage of videos is purely incidental to its core mission of “revolutionizing entertainment.” JAI-855. As such, YouTube’s conduct extends far beyond storage and falls outside the safe harbor. III. Viacom Is Entitled To Summary Judgment On Its Affirmative Claims YouTube’s participation in the rampant infringement on its site not only re- moves it from the safe harbor, but also entitles Viacom to summary judgment on its affirmative claims. YouTube’s unsupported, revisionist history notwithstanding, the core of its conduct is largely undisputed; the only question is whether, as a matter of law, that conduct gives rise to liability. Under established governing principles, it does. A. Viacom Is Entitled To Summary Judgment On Its Direct Infringement Claim YouTube barely contests Viacom’s entitlement to summary judgment on direct infringement, offering on the merits only the assertion that Viacom did not satisfy the “volitional conduct” requirement of Cartoon Network LP v. CSC Holdings, Inc., 536 F.3d 121 (2d Cir. 2008). YT Br. 83 n.37. But in Cartoon Network, the defendant simply supplied a DVR device to users, then played no further role, while “the person who actually presse[d] the button to make the recording . . . supplie[d] the necessary element of volition.” Cartoon Network, 536 F.3d at 131. In contrast, YouTube intentionally chooses to convert the infringing videos into vari- 36 ous formats suitable for a variety of media platforms; to display those videos on demand to its users; and to distribute specific videos to third parties. It promotes and organizes the videos on its site to ensure maximum infringement. As such, YouTube’s conduct satisfies the Cartoon Network standard, and permits entry of summary judgment for Viacom. B. Viacom Is Entitled To Summary Judgment On Its Grokster Claim For Intentional Facilitation Of Infringement Because YouTube is ineligible for the safe harbor, its lead defense against Viacom’s Grokster claim—that “no court has ever held a DMCA-compliant service liable for inducement,” YT Br. 82—misses the point. YouTube suggests that where a service provider makes any effort to combat infringement—no matter how insubstantial—it necessarily lacks intent to induce infringement. But a service provider with intent to promote infringement, like YouTube, also would attempt to, in YouTube’s words, “cover our asses so we don’t get sued,” JAII-112-13, undertaking measures that foster “the perception . . . that we were concerned about this type of material” without “cut[ting] down the actual amount of illegal conduct uploaded.” JAII-176, 183. This pretextual approach to intellectual property enforcement does not negate wrongful intent; it establishes such intent. YouTube attempts to escape Grokster by explaining that even if it “intended to bring about” infringement, it did not “promot[e]” infringement, apparently because it did not “communicate[] an inducing message” to users. YT Br. 84-87. 37 But by remaining idle in the face of widespread publicity regarding the infringing uses of its site, YouTube did communicate a message to its users that it was content to be the “Napster of Video.” JAII-183. In any event, Grokster specifically held that “[p]roving that a message was sent out,” was “not [the] exclusive way of showing that active steps were taken with the purpose of bringing about infringing acts.” Metro-Goldwyn-Meyer Studios Inc. v. Grokster, Ltd., 545 U.S. 913, 938 (2004). Here, YouTube’s disabling of community flagging that would have enabled it to limit such infringement and its purposeful decision to selectively make Audible Magic and Content ID available only to those entering into agreements with it demonstrate such a purpose, and thus “promote” infringement within the meaning of Grokster. Id.9 C. Viacom Is Entitled To Summary Judgment On Its Vicarious Liability Claim YouTube summarily asserts that even under the common law vicarious liability test that it attempts to escape in construing the safe harbor, it neither has the practical ability to control the infringing conduct nor draws a financial benefit from that conduct. YT Br. 83 n.37. Accordingly, YouTube’s argument fails for the reasons addressed above, including its “‘practical ability’” to at least “‘limit the di- 9 That YouTube has non-infringing uses, YT Br. 86-89, does not undermine Viacom’s proof that YouTube also intended to facilitate infringement. 38 rectly infringing conduct’” through digital fingerprint filtering or community flagging. Id. (quoting Perfect 10, 488 F.3d at 1173). YouTube’s protest that copyrighted content—the source of the vast majority of its views—does not act as a draw to users, generating a financial benefit to YouTube, is similarly disproven by YouTube’s own contemporaneous statements. CONCLUSION YouTube’s best efforts cannot obscure the key fact at the core of this case: rather than taking reasonably available steps to reduce or eliminate rampant infringement which it knew was ongoing, YouTube made conscious efforts to avoid obtaining specific knowledge of the infringing clips, and instead “welcomed” that infringement. And it did so for the simple reason that copyrighted works are popular—causing infringement to be profitable. The district court’s conclusion that Congress shielded such intentional infringing conduct through a safe harbor designed for innocent service providers cannot withstand scrutiny. This Court should reverse that judgment and instead grant summary judgment in favor of Viacom. 39 DATED: April 28, 2011 Respectfully submitted, /s/ Theodore B. Olson Paul M. Smith William M. Hohengarten Scott B. Wilkens Matthew S. Hellman JENNER & BLOCK LLP 1099 New York Avenue, NW Washington, DC 20001 (202) 639-6000 Theodore B. Olson Matthew D. McGill GIBSON, DUNN & CRUTCHER LLP 1050 Connecticut Avenue, NW Washington, DC 20036 (202) 955-8500 Susan J. Kohlmann JENNER & BLOCK LLP 919 Third Avenue New York, NY 10022 (212) 891-1600 Stuart J. Baskin SHEARMAN & STERLING LLP 599 Lexington Avenue New York, NY 10022 (212) 848-4000 40 Federal Rules of Appellate Procedure Form 6. Certificate of Compliance With Rule 32(a) Certificate of Compliance With Type-Volume Limitation, Typeface Requirements and Type Style Requirements 1. This brief complies with the type-volume limitation of Fed. R. App. P. 32(a)(7)(B), as modified by this Court’s Order of April 19, 2011 (Document 379), because: X this brief contains 8,938 words, excluding the parts of the brief exempted by Fed. R. App. P. 32(a)(7)(B)(iii), or this brief uses a monospaced typeface and contains [state the number of] lines of text, excluding the parts of the brief ex empted by Fed. R. App. P. 32(a)(7)(B)(iii). 2. This brief complies with the typeface requirements of Fed. R. App. P. 32(a)(5) and the type style requirements of Fed. R. App. P. 32(a)(6) because: X this brief has been prepared in a proportionally spaced typeface using Microsoft Word 2003 in 14 point Times New Roman, or this brief has been prepared in a monospaced typeface using [state name and version of word processing program] with [state number of characters per inch and name of type style]. (s) /s/ Theodore B. Olson Attorney for Plaintiffs-Appellants Dated: April 28, 2011 41 CERTIFICATE OF SERVICE I HEREBY CERTIFY that on this 28th day of April, 2011, a true and correct copy of the foregoing Plaintiffs-Appellants’ Reply Brief was served on all counsel of record in this appeal via CM/ECF pursuant to Local Rule 25.1(h)(1) & (2). /s/ Theodore B. Olson Theodore B. Olson Max W. Berger John C. Browne BERNSTEIN LITOWITZ BERGER & GROSSMANN LLP 1285 Avenue of the Americas New York, New York 10019 (212) 554-1400 Charles S. Sims William M. Hart Noah Siskind Gitterman Elizabeth A. Figueira PROSKAUER ROSE LLP 1585 Broadway New York, New York 10036 (212) 969-3000 Attorneys for Appellants The Football Ass’n Premier League Ltd., et al. (10-3342) Andrew H. Schapiro A. John P. Mancini Brian M. Willen MAYER BROWN LLP 1675 Broadway New York, New York 10019 (212) 506-2500 David H. Kramer Michael H. Rubin Bart E. Volkmer WILSON SONSINI GOODRICH & ROSATI 650 Page Mill Road Palo Alto, California 94304 (650) 493-9300 Attorneys for Appellees 42